hornetball

iTrader: (4)

I have my MS controlling my water injection pump through a relay. The problem I have is flyback voltage. I put in a flyback diode and this mostly remedied the problem. But on rare occasion, I still have some artifacts from the flyback voltage (it sometimes causes a quick MS reset -- it can also cause wierdness on my electronic boost/AFR gauge (dimming or sometimes the gauge freezes).

I'd like to just replace the electromechanical relay with an electronic version. Does anyone have a recommendation for a decent 12V electronic relay?

Thanks!

DIYAutoTune

How is your grounding? this could be an issue.

If you could, do you have a data log of this event occurring? you can e-mail it to me at ForumHelp@diyautotune.com

Reverant

iTrader: (10)

How many amps? Most pumps are 10-11 amps max, but its best to know for sure.

hornetball

iTrader: (4)

For more information, this was the original problem:

https://www.miataturbo.net/megasquir...uirt-wi-59614/

Moving the flyback diodes out to the relays helped a lot, but hasn't completely eliminated the issue. I mostly see it on my VEI gauge (often freezes right at 3psi -- my WI turn-on point). MS resets are rare -- it handles the events much better than the gauge.

Grounding is wired to stock ground points with OEM wire in good condition. Terminals are wire-brushed and clean and nuts are tight. Obviously, when you feed flyback voltage into your ground plane . . . .

I tried to get a data log but haven't captured the issue yet. Like I said, it is now a relatively rare event, but I would like to get rid of it completely, especially the lock-ups on the VEI gauge. I've been compensating by driving in a style that minimizes time around that 3psi turn-on point.

The WI kit is a "Cooling Mist" kit. Uses the common 150psi water pump with a Cooling Mist sticker on it. Sticker recommends a 10-amp slow-blow fuse, which is what I've used.

Reverant

iTrader: (10)

Crydom CMX60D20.

hornetball

iTrader: (4)

Thanks. Ended up getting a Crydom EL100D10-12. Wanted something that would more directly replace the existing eletromechanical relays without needing to do a board. Data sheet:

http://www.mouser.com/ds/2/93/e_l_dc-26114.pdf

Will advise on how this works.

Reverant

iTrader: (10)

10A rating on a pump that will do 10-11A? Think again.

hornetball

iTrader: (4)

I would expect 10A to be the starting current for the pump, not the continuous. After all, you wouldn't protect it with a 10A slow blow fuse if it used 10A all the time. Spec sheet says the relay can handle 100A short term for inrush. I think it will be OK, but if you're right I'll find out soon enough. I was attracted by the form factor (almost a direct replacement for what's there now).

hornetball

iTrader: (4)

Just to wrap this up . . .:

Swapped in the electronic relays and everything is working perfectly. Flyback voltage is gone. Really nice. Swap was easy because the form factor was so similar to the electromechanical relay that was replaced.

Before putting the relays in, I put an ammeter on the pump. Steady state is ~1.5 amps on the battery (would be less with alternator voltage). So, 10 amps would be like an absolute worst case startup current.

JasonC SBB

A cheaper solution would have been to mount the flyback diode on the relay itself.

hornetball

iTrader: (4)

Been there. Done that. Wasn't working for me.

Even with a flyback diode, the ground plane still comes up to Vcc + the diode biasing voltage. A flyback diode will keep the smoke inside the little black bits, but is no guarantee against processor resets, lock-ups, etc.

JasonC SBB

I don't believe you couldn't have fixed it without the e-relays. Flyback voltage don't "inject voltage into the ground plane" nor the "ground plane come up to Vcc". When you put a clamp diode the flyback voltage on the relay is clamped to <15V (if it's a pulldown output driving the relay). If the diode is mounted on the relay the only noisy bit is the wire going to the relay. Worst case than you add an RC "snubber" on that line. And this would only be a problem depending on how fast the MS switches that line.

Can Reverant tell me the value of the gate resistor feeding the MOSFET that powers said relay?

I presume your relay is driven by a pulldown ECU output?

How is the pump wired? Perhaps the noise is coming from the pump.

hornetball

iTrader: (4)

Circuit was posted in https://www.miataturbo.net/megasquir...uirt-wi-59614/. Basically, just the WI control circuit out of the MSnS-Extra Hardware Manual. Relay is activated with a regular old 2N2222 transistor.

Reverant was just helping out with a recommendation. It's an old MSPNP -- not something that Dmitris was supplying.

At this point, the new relay is in and the problem is resolved. It was worth $35 to me to resolve this once and for all after living with it for 15K miles.